Attack categories ranked by interactions
This page ranks the attack categories this site tests by the interactions the live honeypot network records under each one. Visits to a category's attack pages, fetches of its canary URLs and report posts filed under it all count as interactions. A callback is a fetch of the canary URL carried in a payload, not evidence that an agent acted on it. Where attribution exists it resolves callbacks to browsers, crawlers, and scripted HTTP clients, not to agents acting on an instruction, so read this as a ranking of recorded traffic, not proof of compromise.
Interactions by category
Higher bars = more interactions recorded for that category. Click any category to open its first attack page.
Injection classes observed in the wild
Independent signal from HoneyMap, our distributed surface scanner. These are real injection attempts discovered across 240 domains and 294 distinct surfaces.
Where injections actually hide
Surface types that carry real injections in the wild, ranked by count. Every one of these is fair game for an agent reading the page.
- hidden_text154
- html_comment59
- script_literal49
- alt_aria17
- meta_tag8
- data_attr7
How we count
A callback is a fetch of the canary URL carried in a payload, not evidence that an agent acted on it. It establishes that the payload was retrieved and the embedded URL was requested. It does not establish who requested it or why: a browser prefetching every link on a page produces the same record as an agent obeying the instruction, and attribution says most of these are the former.
A report post is a POST to /api/report carrying an attack ID, the confirmation step the planted instructions ask for, and is recorded apart from callbacks. It is a separate and much rarer event, and a deeper signal because it requires acting on the payload rather than fetching a URL from it. It is shown as report posts on the stats page. In this ranking a report post counts as one interaction under its category, as a page visit or a canary URL fetch does. In the registry's JSON, canary URL fetches are the registry field payloadCallbacks and report posts are the registry field canaryTriggers.
Categories on this page are read from the live Registry, and the ranking changes as new interactions arrive. Only the 11 attack categories this site tests are ranked or counted in "Categories tested". Interactions recorded under any other label are left out of the ranking, including visits to trap, tool and research-article pages and report posts the site could not file under one of the 11. The Total interactions figure on this page still includes them.
Test your agent against all of them
HackMyAgent scans the agent setup in the directory you run it from, a static read of its files, and prints each finding with its severity. It does not run your agent. No accounts, no setup, takes a few seconds on a small repository.
npx hackmyagent secure